GPT-5 just cracked Sudoku

PLUS: AI's 'Godmother' on spatial intelligence and Google's new PDF-to-video tool

Author

Gen AI Art
November 12, 2025 • Estimated Reading Time: 9 minutes

Together with

GPT-5 has achieved a new milestone in AI reasoning, becoming the first model to successfully solve a complete 9x9 Sudoku puzzle. This marks a notable advance in teaching AI to handle complex, rule-based logical problems.

The model's 33% success rate on the benchmark more than doubles previous efforts, yet a majority of puzzles remain unsolved. Does this breakthrough signal genuine progress in deep reasoning, or does it mainly highlight the gap between AI's logic and human creativity?

Today in AI:

  • GPT-5 becomes first AI to solve Sudoku

  • AI's 'Godmother' on spatial intelligence

  • Google's new PDF-to-video tool

PRESENTED BY SKEJ

Meet your new assistant (who happens to be AI).

Meet Skej — your new scheduling assistant. Whether it’s a coffee intro, a client check-in, or a last-minute reschedule, Skej is on it. Just CC Skej on your emails, and it takes care of everything: checking calendars, suggesting times, and sending out invites.

What’s new? In a significant benchmark for deep reasoning, GPT-5 just became the first AI to solve a full 9x9 Sudoku puzzle.

What matters?

  • The achievement was recorded on Sakana AI’s Sudoku-Bench, a test specifically designed to push AI models beyond simple logic into multi-step spatial reasoning.

  • GPT-5 achieved a 33% solve rate across the benchmark's puzzles, roughly doubling the performance of the previous leading model.

  • Despite the leap, 67% of the puzzles remain unsolved, highlighting current AI struggles with creative problem-solving and learning novel rules on the fly.

Why it matters?

This breakthrough demonstrates real progress in teaching models to handle structured, rule-based logic puzzles. However, the unsolved puzzles show the remaining gap between an AI's computational power and the creative, intuitive insight humans use to solve new problems.

GUIDE

What’s new? Dr. Fei-Fei Li, a key figure in modern AI, argues in a new essay that language models have hit a wall and the next breakthrough requires 'spatial intelligence'—the ability for AI to understand and interact with the 3D world.What matters?

  • Today’s AI can write code and explain complex topics but is fundamentally blind to the physical world, like a genius who has never left their study.

  • Li proposes that developing this ability requires world models, a new type of AI that can generate virtual worlds, process multimodal data like video and actions, and predict physical outcomes.

  • This leap unlocks future applications in robotics and scientific simulation, enabling AI to solve tangible problems by understanding how objects move and interact in space.

Why it matters?

This shift could mean the difference between AI that only describes the world and AI that truly understands and operates within it. Moving beyond language alone will allow AI to tackle physical challenges, from navigating complex environments to accelerating scientific discovery.

PRESENTED BY RAD INTEL

$1K Could’ve Made $2.5M

In 1999, $1K in Nvidia’s IPO would be worth $2.5M today. Now another early-stage AI tech startup is breaking through—and it’s still early.

RAD Intel’s award-winning AI platform helps Fortune 1000 brands predict ad performance before they spend.

The company’s valuation has surged 4900% in four years* with over $50M raised.

Already trusted by a who’s-who roster of Fortune 1000 brands and leading global agencies. Recurring seven-figure partnerships in place and their Nasdaq ticker is reserved: $RADI.

Now open at $0.81/share, allocations limited – price moves on 11/20.

This is a paid advertisement for RAD Intel made pursuant to Regulation A+ offering and involves risk, including the possible loss of principal. The valuation is set by the Company and there is currently no public market for the Company's Common Stock. Nasdaq ticker “RADI” has been reserved by RAD Intel and any potential listing is subject to future regulatory approval and market conditions. Investor references reflect factual individual or institutional participation and do not imply endorsement or sponsorship by the referenced companies. Please read the offering circular and related risks at invest.radintel.ai.

What’s new? Google's NotebookLM now includes a 'Video Overview' feature, which automatically turns documents, PDFs, and notes into shareable explainer videos with narration and visuals.

What matters?

  • It can synthesize up to 50 different sources—including PDFs, Google Docs, and YouTube transcripts—into a single video summary.

  • Users can direct the AI to focus on specific topics, adjust the tone for different audiences, and even learn advanced techniques for better results.

  • The tool generates a downloadable video in minutes, enabling rapid content creation for training modules, client reports, and marketing briefs.

Why it matters?

This feature drastically lowers the barrier for creating video content, turning dense information into easily digestible visual summaries. It marks another step toward AI assistants that not only organize information but also transform it into new, more accessible formats.

Everything else in AI

Anthropic projects a major cost advantage over OpenAI, expecting to spend far less on compute for its AI models over the next few years thanks to a more diverse chip strategy using hardware from Amazon, Google, and Nvidia.

Microsoft enabled its Copilot desktop app with Voice and Vision features, allowing users to analyze spreadsheet data hands-free by asking questions aloud and getting instant insights without typing formulas.

NPR covered an unofficial movement to “save the em-dash” from its perceived overuse in AI-generated text, highlighting a push to preserve the punctuation's human-driven stylistic nuance.

Essential AI Guides - Reading List:

Let us know!

Work with us

Reach 100k+ engaged Tech Professionals, Engineers, Managers and decision makers. Join brands like MorningBrew, HubSpot, Prezi, Nike, Ahref, Roku, 1440, Superhuman, and others in showcasing your product to our audience. Get in touch now →